potentiation of trpm7 inward currents by...

The

Jour

nal o

f G

ener

al P

hysi

olo

gy

J. Gen. Physiol. © The Rockefeller University Press $8.00Volume 126 Number 2 August 2005 137–150http://www.jgp.org/cgi/doi/10.1085/jgp.200409185

137

ARTICLE

Potentiation of TRPM7 Inward Currents by Protons

Jianmin Jiang, Mingjiang Li, and Lixia Yue

Center for Cardiology and Cardiovascular Biology, Department of Cell Biology, University of Connecticut Health Center, Farmington, CT 06032

TRPM7 is unique in being both an ion channel and a protein kinase. It conducts a large outward current at

�

100 mV but a small inward current at voltages ranging from

�

100 to

�

40 mV under physiological ionic conditions.Here we show that the small inward current of TRPM7 was dramatically enhanced by a decrease in extracellularpH, with an

�

10-fold increase at pH 4.0 and 1–2-fold increase at pH 6.0. Several lines of evidence suggest thatprotons enhance TRPM7 inward currents by competing with Ca

2

�

and Mg

2

�

for binding sites, thereby releasingblockade of divalent cations on inward monovalent currents. First, extracellular protons significantly increasedmonovalent cation permeability. Second, higher proton concentrations were required to induce 50% of maximalincrease in TRPM7 currents when the external Ca

2

�

and Mg

2

�

concentrations were increased. Third, the apparentaffinity for Ca

2

�

and Mg

2

�

was significantly diminished at elevated external H

�

concentrations. Fourth, theanomalous-mole fraction behavior of H

�

permeation further suggests that protons compete with divalent cationsfor binding sites in the TRPM7 pore. Taken together, it appears that at physiological pH (7.4), Ca

2

�

and Mg

2

�

bind to TRPM7 and inhibit the monovalent cationic currents; whereas at high H

�

concentrations, the affinity ofTRPM7 for Ca

2

�

and Mg

2

�

is decreased, thereby allowing monovalent cations to pass through TRPM7. Furthermore,we showed that the endogenous TRPM7-like current, which is known as Mg

2

�

-inhibitable cation current (MIC) orMg nucleotide–regulated metal ion current (MagNuM) in rat basophilic leukemia (RBL) cells was also significantlypotentiated by acidic pH, suggesting that MIC/MagNuM is encoded by TRPM7. The pH sensitivity represents anovel feature of TRPM7 and implies that TRPM7 may play a role under acidic pathological conditions.

I N T R O D U C T I O N

TRPM7 is a ubiquitously distributed ion channel thatbelongs to the long or melastatin-related transientreceptor potential (TRPM) ion channel subfamily(Harteneck et al., 2000; Montell, 2001; Clapham, 2003;Fleig and Penner, 2004). It is unique in being both anion channel and a protein kinase. Although the physio-logical functions of the kinase are not well understood,recent studies have suggested that TRPM7 plays impor-tant roles in cellular Mg

2

�

homeostasis (Schmitz et al.,2003), anoxic neuronal cell death (Aarts et al., 2003),cell proliferation and viability (Nadler et al., 2001;Hanano et al., 2004), and diseases caused by abnormalmagnesium absorption (Schlingmann et al., 2002; Wal-der et al., 2002; Chubanov et al., 2004).

TRPM7 produces pronounced outward currents atnonphysiological voltages ranging from

�

50 to

�

100 mVand small inward currents at negative potentials between

�

100 to

�

40 mV when expressed heterologously inmammalian cells (Nadler et al., 2001; Runnels et al.,2001; Monteilh-Zoller et al., 2003; Schmitz et al., 2003).Unlike some other TRP channels that are gated orpotentiated by activation of the PLC pathway (Clap-ham, 2003), TRPM7 is inhibited by depletion of PIP

2

mediated by PLC activation (Runnels et al., 2002;Aarts et al., 2003). The basal activity of TRPM7 is regu-lated by millimolar levels of intracellular MgATP andMg

2

�

, so that TRPM7 is activated by depletion of intra-cellular MgATP and Mg

2

�

, and is inhibited by high con-centrations of MgATP and Mg

2

�

with an IC

50

of

�

0.6mM (Nadler et al., 2001). The mechanism by whichMg

2

�

inhibits TRPM7, however, is not yet entirely clear(Nadler et al., 2001; Hermosura et al., 2002; Prakriya andLewis, 2002; Runnels et al., 2002; Kerschbaum et al.,2003; Kozak and Cahalan, 2003; Monteilh-Zoller et al.,2003; Schmitz et al., 2003). Other divalent cationshave also been reported to inhibit TRPM7 (Kozak andCahalan, 2003).

Although inactivation of TRPM7 has been extensivelyinvestigated, the activation mechanism of TRPM7 un-der physiological conditions remains unknown. Intra-cellular Mg

2

�

levels (0.5–1 mM) under physiologicalconditions can inactivate

�

50% of TRPM7 channelactivities (Nadler et al., 2001; Kozak and Cahalan,2003). Thus, the inward current amplitude, which isusually 1/30 to 1/10 of the outward current amplitudemeasured at

�

100 mV (Nadler et al., 2001; Runnels etal., 2001; Schmitz et al., 2003), may be very small under

J. Jiang and M. Li contributed equally to this work.Correspondence to Lixia Yue: [email protected]. Li’s present address is Renmin Hospital of Wuhan University,

People’s Republic of China.

Abbreviations used in this paper: DVF, divalent-free solution; MagNuM,Mg nucleotide–regulated metal ion current; MIC, Mg

2

�

-inhibitablecation; RBL, rat basophilic leukemia; TRPM, melastatin-relatedtransient receptor potential.

138 Protons Increase Monovalent Permeability of TRPM7

physiological internal Mg

2

�

levels in the native cells.Given the potential important physiological functions(Nadler et al., 2001; Runnels et al., 2002; Schlingmannet al., 2002; Walder et al., 2002; Aarts et al., 2003; Ryaza-nova et al., 2004; Schmitz et al., 2003; Chubanov et al.,2004; Hanano et al., 2004), it is likely that TRPM7inward current may be potentiated by physiologicalor pathological stimuli. A recent study showed thatTRPM7 is up- and down-regulated in a cAMP- and PKA-dependent manner (Takezawa et al., 2004), with thechanges assessed by outward current amplitude. H

2

O

2

was reported to increase TRPM7 inward currents byone- to twofold after prolonged (30–50 min) incuba-tion (Aarts et al., 2003). However, the mechanism bywhich H

2

O

2

regulates TRPM7 remains unclear.In the present study, we demonstrate that protons

markedly potentiate TRPM7 inward currents. Loweringextracellular pH increases the inward current by

�

10-fold, whereas the outward current is only changed by

�

30%. This transforms the normal outward rectifica-tion of TRPM7 to rectification in both the inward andoutward directions. Further, our data suggest that pro-tons enhance TRPM7 inward currents by competingwith external Ca

2

�

and Mg

2

�

for binding sites in theTRPM7 pore. Therefore, the effect of acidic pH onTRPM7 is more pronounced when extracellular Ca

2

�

concentration is decreased. Importantly, we show thatthe well-characterized endogenous TRPM7-like currentMIC (Mg

2

�

-inhibitable cation)/MagNuM (Mg nucle-otide–regulated metal ion) in rat basophilic leukemia(RBL) cells is similarly potentiated by a decrease in ex-tracellular pH. As high proton concentrations (pH

�

6) can be generated during various forms of injury, in-cluding infection, inflammation, and ischemia (Jaco-bus et al., 1977; Stevens et al., 1991; Steen et al., 1992),the significant increase in TRPM7 inward currents byprotons suggests that TRPM7 may play a role underacidic pathological conditions.

M A T E R I A L S A N D M E T H O D S

Cell Culture

HEK-293 cells stably transfected with a FLAG-tagged murineTRPM7 in pCDNA4/TO vector were provided by A. Scharenberg(University of Washington, Seattle, WA). Cells were grown withDMEM medium supplemented with 10% FBS, 100 U/ml penicil-lin and 100 mg/ml streptomycin, blasticidin (5

�

g/ml), and zeo-cin (0.4 mg/ml). Expression of TRPM7 was induced 12–24 h be-fore experiments by adding 1

�

g/ml tetracycline to the culturemedium (Nadler et al., 2001; Schmitz et al., 2003). Unless other-wise stated, experiments were conducted using TRPM7 express-ing HEK-293 cells after tetracycline induction.

RBL-2H3 cells were provided by D. Clapham (Harvard Medi-cal School, Boston, MA). Cells were cultured in DMEM supple-mented with 10% FBS and 100 U/ml penicillin and 100 mg/mlstreptomycin. For electrophysiological experiments, cells wereplated onto glass coverslips and used 12 h thereafter.

Electrophysiology

Whole-cell currents were recorded using an Axopatch 200B(Axon Instruments, Inc.) amplifier. Data were digitized at 20 kHzand low-pass filtered at 2 kHz. pClamp9 software was used fordata acquisition and analysis. Patch electrodes were pulled fromborosilicate glass and fire polished to a resistance of

�

3 M

�

when filled with internal solutions. Series resistance (R

s

) wascompensated up to 90% to reduce series resistance errors to

�

5mV. Cells in which R

s

was

�

10 M

�

were discarded (Yue et al.,2002). All experiments were conducted at 22

�

2

�

C.Voltage stimuli lasting 250 ms were delivered at 1- to 5-s inter-

vals, with either voltage ramps or voltage steps ranging from

�

120to

�

100 mV. Unless otherwise stated, 3–5 min were allowed to letTRPM7 current develop and reach a steady state after break-in. Afast perfusion system was used to exchange extracellular solu-tions. A complete solution exchange can be achieved in

�

1–3 s.The internal pipette solution (P1) for whole cell current re-

cordings in HEK-293 cells stably expressing TRPM7 contained(in mM) 145 Cs-methanesulfonate, 8 NaCl, 10 EGTA, and 10HEPES, pH adjusted to 7.2 with CsOH. In some experiments,Mg

2

�

was added to the pipette solution and the free Mg

2

�

con-centration was titrated to 3 mM (calculated with the MaxChela-tor software, available at http://www.stanford.edu/~cpatton/webmaxcS.htm). For current recordings in RBL cells, pipette so-lution (P2) contained (in mM) 145 Cs-methanesulfonate, 8NaCl, 1 EGTA, 0.084 Ca

2

�

, 1 MgCl

2

, 5 ATP-Na

2

, 10 HEPES, pHadjusted to 7.2 with CsOH. Free Ca

2

�

and Mg

2

�

concentrationswere estimated at

�

15 nM and 26

�

M, respectively (MaxChela-tor). In experiments designed to diminish outward currents, pi-pette solution (P3) contained (in mM) 120 NMDG, 108 glutamicacid, 10 HEPES, 10 EGTA, 10 CsCl, pH adjusted 7.2 with NMDG.

The standard extracellular Tyrode’s solution contained (inmM) 140 NaCl, 5 KCl, 2 CaCl

2

, 20 HEPES, and 10 glucose, pH ad-justed to 7.4 (NaOH). HEPES (20 mM) was used in the solutionsat pH 7.0 and 7.4, and was replaced by 10 mM HEPES and 10 mMMES for the solutions at pH

6 (Jordt et al., 2000; Askwith et al.,2004; Yermolaieva et al., 2004). Divalent-free solution (DVF) con-tained (in mM) 145 NaCl, 20 HEPES, 5 EGTA, 2 EDTA and 10 glu-cose, with estimated free [Ca

2

�

]

�

1 nM at pH 7.4 and free[Mg

2

�

]

�

10 nM at pH 7.4 (MaxChelator). HEPES (20 mM) wasreplaced by 10 mM HEPES and 10 mM MES in DVF solutions atpH 4.0, and the estimated free [Ca

2

�

] was 7.7

�

M and free [Mg

2

�

]was 9.9

�

M in DVF at pH 4.0 (MaxChelator). Appropriate Ca

2

�

orMg

2

�

was added to the DVF at pH 7.4 to prepare solutions contain-ing

10

�

M Mg

2

�

or Ca

2

�

(Fig. 6). Solutions containing 1, 2, and10 mM Mg

2

�

or Ca

2

�

at both pH 4.0 and 7.4 were prepared byomitting EDTA and EGTA in the DVF solution, and by adding theappropriate concentrations of Mg

2

�

or Ca

2

�

. Isotonic Ca

2

�

or Mg

2

�

solution contains 120 mM Ca

2

�

or Mg

2

�

, 10 mM HEPES, 10 mMglucose, with pH adjusted to pH 7.4 or pH 4.0. Anomalous molefraction behavior of Ca

2

�

permeation (Fig. 3 F) was evaluated in aseries of external solutions, including isotonic Ca

2

�

(120 mM), 10mM Ca

2

�

, 2 mM Ca

2

�

, 1 mM Ca

2

�

, 100

�

M Ca

2

�

, and nominallyCa

2

�

-free solution in which free Ca

2

�

concentration was estimatedat 10

�

M (Vennekens et al., 2000; Yue et al., 2001). The solutionscontaining 100

�

M to 10 mM Ca

2

�

were prepared from normal Ty-rode’s solution by adding appropriate concentration of Ca

2

�

, withreductions in Na

�

concentration when necessary to keep the con-stant osmolarity. The same method was used to prepare a series ofsolutions containing various Mg

2

�

concentrations for anomalousmole fraction behavior of Mg

2

�

permeation experiment shown inFig. 3 F. Cells were usually exposed to acidic solutions for

�

30 s toavoid desensitization unless otherwise stated. Current amplitudewas measured at

�

120 or

�

100 mV. Amiloride was added to theperfusate as indicated in the text. All the chemicals for electro-physiological experiments were from Sigma-Aldrich.

Jiang et al. 139

Data AnalysisPooled data are presented as mean � SEM. Concentration–response curves were fitted by an equation of the form: E Emax{1/[1 � (EC50/C)n]}, where E is the effect at concentration C,Emax is maximal effect, EC50 is the concentration for half-maximaleffect and n is the Hill coefficient (Yue et al., 2000). EC50 is re-placed by IC50 if the effect is an inhibitory effect. Statistical com-parisons were made using two-way analysis of variance (ANOVA)and two-tailed t test with Bonferroni correction; P � 0.05 indi-cated statistical significance.

R E S U L T S

Potentiation of TRPM7 Inward Currents by ProtonsTRPM7 currents were elicited by voltage ramps or volt-age steps ranging from �120 to �100 mV from a hold-ing potential of 0 mV (Runnels et al., 2001). As previ-ously reported, TRPM7 produced large outward cur-rents and small inward currents (Fig. 1 A, red trace)(Nadler et al., 2001; Runnels et al., 2001; Schmitz et al.,2003). After break-in, 3–5 min was allowed to letTRPM7 current amplitude reach a steady state (seeMATERIALS AND METHODS) before changing exter-nal solutions. As the measurable outward current oc-curs at nonphysiological range, we investigatedwhether the TRPM7 inward current can be potentiatedby pathological stimuli. As shown in Fig. 1 A, the in-ward current of TRPM7 was dramatically increased by adecrease in the extracellular pH to 4.0. In the same cellas shown in Fig. 1 A, a similar degree of increase in in-ward current was observed upon a second applicationof the external solution at pH 4.0, indicating that theeffect of pH 4.0 on TRPM7 was reversible and repro-ducible (Fig. 1 B). While the inward currents were in-creased by �10-fold (at �120 mV) at pH 4.0, the out-ward currents measured at �100 mV were only mildlychanged, showing an �30% increase (Fig. 1 B, top).The large increase in inward current compared withthe small change in outward current induced by acidicpH was also evident in the currents elicited by voltagesteps ranging from �120 to �100 mV (Fig. 1, C and D).

Figure 1. Potentiation of TRPM7 inward currents by externalprotons in HEK-293 cells stably expressing TRPM7. (A) Represen-tative TRPM7 currents evoked by voltage ramps ranging from�120 to �100 mV in the external Tyrode’s solutions at pH 7.4 andpH 4.0. (B) Normalized outward (�100 mV) and inward currents(�120 mV) at pH 4.0 and pH 7.4. The acidic external solution(pH 4.0) was applied after the current amplitude reached a steady-state (�5 min after break-in). Repetitive application of acidicsolution produced a similar increase in inward and outwardcurrents. (C and D) TRPM7 currents elicited by voltage stepsranging from �120 to �100 mV with an increment of 20 mV atpH 7.4 (C) and pH 4.0 (D). Only inward current was significantlyincreased at pH 4.0 (D). (E) Currents recorded at various timepoints in a cell dialyzed with the pipette solution containing 3 mMfree Mg2�. a, immediately after formation of whole-cell configura-tion; b, before the external solution was changed from pH 7.4 topH 4.0; c, the first time application of external solution at pH 4.0induced an increase in inward current; d, the fifth time application

of external solution at pH 4.0 could not induce any changebecause TRPM7 was completely blocked by intracellular Mg2�. (F)In the same cell as shown in E, continuous changes in inward(top) and outward (middle) currents measured at �100 mV and�120 mV were plotted as a function of time. The inward currentslabeled 1, 2, 3, 4, 5, and 6 (middle) represent the time points whenthe cell was exposed to the acidic external solution (pH 4.0).(Bottom) Normalized inward current at pH 4.0 (filled circle, red)superimposed with the normalized outward current at pH 7.4(filled triangle, black). The green line obtained by fitting thenormalized outward current represents currents decay. (G)TRPM7 currents recorded at pH 7.4 and pH 4.0 with and without200 �M amiloride (Ami), respectively. (H) Time-dependentchanges of the inward current induced by pH 4.0 in the presenceand absence of 200 �M amiloride (Ami).


To test whether the changes elicited by low pH weremediated by TRPM7, we studied the effects of acidic pHsolution on control cells. In HEK-293 cells without tetra-cycline induction of TRPM7 expression, small endoge-nous TRPM7-like currents were recorded. The endoge-nous inward currents were also enhanced by �10-fold atpH 4.0 (n 6; unpublished data). Although the out-wardly rectifying I-V curve suggests that the endogenouscurrent in the HEK-293 cells is TRPM7-like current, wecannot exclude the possibility that the current may befrom some leaky expression. To further confirm thatthe pH 4.0–elicited increases in inward current arethrough TRPM7, we did the following experiments. AsTRPM7 is ubiquitously expressed in various cell types(Nadler et al., 2001; Hermosura et al., 2002; Kozak etal., 2002; Prakriya and Lewis, 2002; Runnels et al., 2002;Aarts et al., 2003; Jiang et al., 2003; Kozak and Cahalan,2003), including the HEK-293 cells used to create thestable TRPM7 expressing cell lines (Nadler et al., 2001;Takezawa et al., 2004), it is difficult to find a cell typethat does not have endogenous TRPM7 expression.Therefore, we used a pipette solution containing 3 mMfree Mg2� to suppress TRPM7 currents and then mea-sured the response of the cells to acidic pH. TRPM7 cur-rents were apparent immediately after formation of thewhole-cell configuration (Fig. 1 E, a and b), but thenthe current amplitude gradually decreased (Fig. 1 F).Application of external solution at pH 4.0 induced dra-matic increases in inward current with a small increasein outward currents early after forming the whole cellconfiguration (Fig. 1 E, c, and Fig. 1 F, 1(c)). Before thecurrent was completely inhibited by intracellular Mg2�,exposing the cell to the external solution at pH 4.0 for asecond, third, and fourth time repeatedly induced sig-nificant increases in TRPM7 inward currents (Fig. 1 F, 1,2, 3, and 4). After TRPM7 was totally blocked by intra-cellular Mg2�, pH 4.0 failed to induce any change in cur-rent amplitude (Fig. 1 E, d, and Fig 1 F, 5 and 6). Simi-lar results were observed in five other cells, suggestingthat pH 4.0–elicited increases in current amplitude arethrough TRPM7. In addition, the normalized inwardcurrent at pH 4.0 plotted as a function of time superim-posed with that of normalized outward current at pH7.4, indicating that the inward current at pH 4.0 and theoutward current at pH 7.4 decay at the same rate whenthe pipette solution contains 3 mM free Mg2� (Fig. 1 F,bottom), further indicating that the pH 4.0–induced in-crease in inward current is mediated by TRPM7.

To rule out the possibility that proton-activated Nachannels (Waldmann and Lazdunski, 1998) were in-volved in the increased inward current elicited by lowpH, amiloride (200 �M) was added to the external so-lution. No significant difference was observed in in-ward currents elicited by pH 4.0 in the presence andabsence of 200 �M amiloride (Fig. 1, G and H), sug-

gesting that the pH 4.0–induced inward currents werenot due to acid-sensitive channels of the degenerinfamily (Waldmann and Lazdunski, 1998). Similarly,contamination by a proton-activated Cl� current (Chernyet al., 1997) was excluded because acidic pH–inducedincreases in TRPM7 currents were not affected by re-placement of NaCl with NaSO3CH3 in the external so-lution (see Fig. 2). All the above-mentioned results sug-gest that the marked increases in inward currentsevoked by acidic pH in TRPM7-overexpressing cellswere mediated by TRPM7 channels.

Figure 2. Concentration-dependent effects of protons on TRPM7currents. (A–E) Potentiation of TRPM7 inward currents by protonsat pH 7.0 (A), pH 6.0 (B), pH 5.0 (C), pH 4.0 (D), and pH 3.0 (E).Representative traces were elicited by voltage ramps ranging from�120 to �100 mV in the external solutions with NaCl replacedby NaSO3CH3. The y axis in each panel was scaled to illustratethe changes in inward currents. (F) Concentration-dependentchanges in inward current amplitude measured at �120 mV fromthe recordings elicited by voltage ramps at indicated pH. Thedashed line represents zero current levels. (G) The increase incurrent amplitude at the indicated pH was normalized to the valueat pH 3.0. Best fit of the normalized data yielded an EC50 4.5 �0.5 pH unit (mean � SEM, n 8; Hill coefficient was 1).

Jiang et al. 141

We next studied concentration-dependent effects ofprotons on TRPM7. A small increase in the inward cur-rent was seen at pH 7.0 (Fig. 2 A); the increase was sig-nificant at pH 6.0 and reached a maximum at pH 3.0(Fig. 2, B–E). The concentration-dependent increaseof TRPM7 inward currents from the same cell is shownin Fig. 2 F. At pH 3.0, the average inward current ampli-tude measured at �120 mV was increased by 10.2 �1.3–fold compared with the current amplitude at pH7.4. The pH required for inducing half-maximal in-crease in inward current was 4.5 (Fig. 2 G). Changes in

the outward current (measured at �100 mV) weremuch smaller than those of the inward current. For ex-ample, 13.1 � 1.1%, 23.1 � 2.5%, 23.3 � 3.1%, and24.8 � 3.4% increases were observed at pH 7.0, 6.0, 5.0,and 4.0 (P � 0.05), respectively; whereas a 26.5 � 4.0%(P � 0.05) decrease was seen at pH 3.0.

Protons Increase Monovalent Cation Permeability through TRPM7We previously reported that TRPM7 is a nonselectivecation channel that is permeable to both monovalent

Figure 3. Monovalent per-meability of TRPM7 at pH7.4. (A and B) Original re-cordings of TRPM7 currentsobtained using P3 pipettesolution in the following ex-ternal solutions at pH 7.4(mM): a, 2 Ca2�/NMDG; b,isotonic NMDG; c, 2 Ca2�/1Mg2�/NMDG; d, 2 Ca2�/1Mg2�/150 Na�; e, 2 Ca2�/1Mg2�/150 K�; f, 2 Ca2�/1Mg2�/150 Cs�; g, isotonicMg2� (120 mM); h, isotonicCa2� (120 mM). The insetsshow an expanded sectionof the graph to illustratechanges in reversal potentialin different solutions. (C) In-ward current amplitude in arepresentative cell in the indi-cated external solutions. (D)Changes in reversal potentialobtained by subtracting re-versal potential in isotonicNMDG from those in test so-lutions (mean � SEM, n 6).**, P � 0.01 in comparisonwith the value obtained in 2Ca2�/1 Mg2�/NMDG. (E)Averaged TRPM7 currentamplitude in various test so-lutions. Note that the currentamplitude in 2 Ca2�/1 Mg2�/150 Na�, 2 Ca2�/1 Mg2�/150K�, and 2 Ca2�/1 Mg2�/150Cs� is significantly larger thanthat in 2 Ca2�/1 Mg2�/NMDG. (F) Ca2�- and Mg2�-dependent anomalous mole-fraction effects. In the leftpanel, changes in inwardcurrent measured at �120mV were normalized in refer-ence to the value at 10 mMextracellular Ca2� (extracellu-lar solutions were nominallyMg2� free). In the right

panel, changes in inward current measured at �120 mV were normalized in reference to the value at 2 mM extracellular Mg2� (extracellularsolutions were nominally Ca2� free). Note that the current amplitude is minimal in the external solutions with 10 mM Ca2� (left) or2 mM Mg2� (right) (mean � SEM, n 6).


and divalent cations (Runnels et al., 2001). Monovalentpermeability in the presence of divalent cations wasalso reported for the MIC channel in Jurkat cells(Prakriya and Lewis, 2002). Other studies suggestedthat TRPM7 is selective for divalent ions, and the in-ward current was carried exclusively by divalent cationsin external solutions containing 10 mM Ca2� and 2 mMMg2� (Nadler et al., 2001). To test whether monovalentcations contribute to TRPM7 inward currents underphysiological Ca2� (2 mM) and Mg2� (1 mM) concen-trations, we compared the TRPM7 current amplitudeand reversal potentials in extracellular solutions con-

taining different monovalent cations, with those in theisotonic divalent cation solutions (Fig. 3 B). The P3 pi-pette solution containing lowered Cs� concentration(10 mM, see MATERIALS AND METHODS) was usedto minimize outward current amplitude (Fig. 3). In thepresence of 2 mM Ca2� and 1 mM Mg2�, changing theexternal solution from nonpermeable NMDG (Jiang etal., 2003; Kozak and Cahalan, 2003) to the solutionscontaining 150 mM Na�, K�, or Cs� significantly in-creased inward current amplitude and shifted rever-sal potentials (Fig. 3, A, C, D, and E), indicating thatunder physiological divalent cation concentrations,

Figure 4. Protons enhance TRPM7 inwardcurrents by increasing monovalent cationpermeability. (A, C, and E) Representativerecordings of TRPM7 in 0, 10, 30, 100, and150 mM Na� (A), K� (C), and Cs� (E) in thepresence of 10 mM Ca2� at pH 4.0, respec-tively. The external solution containing 10mM Ca2�/150 mM NMDG (pH 4.0) was usedas a control solution. When monovalent ionconcentrations were increased, equal molarNMDG was omitted from the solution tokeep the osmolarity constant. Note the con-centration-dependent increase of TRPM7conductance in A, C, and E. The insets showexpanded section of the graphs to illustratechanges in reversal potential in various mono-valent ion concentrations. (B, D, and F) Aver-age changes in reversal potential (top) andinward current density (bottom) measuredat �120 mV in the external solutions withvarious concentrations of monovalent cations.The changes in reversal potential were ob-tained by subtracting reversal potential in thecontrol solution (10 mM Ca2�/NMDG) fromthose in test solutions with various monovalentcation concentrations (mean � SEM; n 6[B]; n 7 [D]; n 9 [F]). *, P � 0.05; **,P � 0.01 in comparison with Erev in 10 mMCa2�/NMDG.

Jiang et al. 143

monovalent cationic currents contribute to TRPM7 in-ward currents. We further studied the relative effects ofdivalent cations on TRPM7 monovalent currents. Fig. 3F shows anomalous mole fraction behavior of Mg2� andCa2� permeation (Fig. 3 F). The smallest current ampli-tude was observed in the external solutions containing10 mM Ca2� or 2 mM Mg2�, respectively.

We proceeded to study the monovalent conductanceunder acidic conditions by using external solutionscontaining different concentrations of Na�, K�, or Cs�

in the presence of 10 mM Ca2� (Fig. 4). The P3 pipettesolution was used in this experiment. Whole-cell con-figuration was established in the 2 mM Ca2� Tyrode’ssolution at pH 7.4, and cells were then exposed to 0,10, 30, 100, and 150 mM Na�, K�, or Cs� solutions inthe presence of 10 mM Ca2� at pH 4.0 for 30–60 s. Con-centration-dependent increases in current amplitudeand shift of reversal potentials at high monovalent cat-ion concentrations at pH 4.0 were observed (Fig. 4, A,C, and E, also see insets). These results (Fig. 4) indicatethat the proton-evoked increases in TRPM7 inward cur-rents were attributed to the increased Na�, K�, and Cs�

conductance.To ensure that protons enhance monovalent conduc-

tance without increasing divalent conductance, we eval-uated the effects of acidic pH on divalent currents. Atphysiological concentrations of Ca2� (2 mM) and Mg2�

(1 mM) in the presence of NMDG solution, protonssignificantly decreased current amplitude measured at�120 mV (Fig. 5, A and B). This result suggests thatprotons may compete with divalent cations for bindingsites in the pore, so that low concentration of divalentcations were outcompeted by protons, therefore the di-

valent current amplitude at pH 4.0 was smaller thanthat at pH 7.4. In agreement with this notion, we foundthat the isotonic Ca2� and Mg2� currents at pH 4.0 werenot significantly different from those at pH 7.4 (Fig. 5,C and D). Thus, the acidic external solution–inducedincrease in inward currents was mediated by increasingmonovalent cation permeability through TRPM7.

Protons Increase TRPM7 Currents by Competing with Ca2� and Mg2� for Binding SitesTo investigate the mechanism by which protons potenti-ate TRPM7 monovalent inward currents, we studied theeffects of protons on TRPM7 currents in the presenceof various external Ca2� or Mg2� concentrations. At pH4.0, protons induced a maximal increase in inward cur-rent in the external solution containing 0.5 mM Ca2�,but only �30% of maximal response in the external so-lution containing 10 mM Ca2� (Fig. 6 A). The EC50 forprotons was changed from pH 5.1 to 4.5 and 3.4 whenthe external Ca2� concentration was increased from 0.5mM to 2 and 10 mM (Fig. 6 B), respectively. Similarly,pH 3.5 elicited a maximal increase in inward current inthe external solution containing 1 mM Mg2�, but only�50% maximal increase in TRPM7 inward current inthe external solution containing 10 mM Mg2�. The EC50

was changed from pH 4.6 to pH 3.6 when the externalMg2� concentration was increased from 1 to 10 mM(Fig. 6, C and D). The 1.1 and 1.7 pH unit change inEC50 when external Ca2� was changed from 0.5 to 2 and10 mM and 1 pH unit change in EC50 when externalMg2� was increased from 1 to 10 mM indicate thatincreasing external Ca2� or Mg2� concentrations de-creases the affinity of TRPM7 for protons, and suggests

Figure 5. Acidic pH does not increaseTRPM7 divalent conductance. Currentswere elicited by voltage ramps (�120 to�100 mV) with P1 pipette solution. (A)Divalent current amplitude measuredat �120 mV in the external solutionscontaining 2 mM Ca2�/1 mM Mg2�/NMDG at pH 7.4 and pH 4.0, respec-tively. Dash line represents zero currentlevel. (B) Averaged divalent currentdensity of TRPM7 measured at �120mV in the external solutions as shownin A (mean � SEM, n 8). Currentdensity at pH 4.0 was significantlysmaller than that at pH 7.4. (C) Currentamplitude of TRPM7 measured at�120 mV in the isotonic Mg2� andisotonic Ca2� solutions at pH 7.4 andpH 4.0, respectively. Dashed lines rep-resent zero current level. (D) Meancurrent density at �120 mV in theisotonic Mg2� or Ca2� solutions at pH7.4 and pH 4.0, respectively (mean �SEM, n 6).


(n 6), �22.0 � 2.2% (n 6), �24.8 � 3.4% (n 8),and �26.2 � 8.9% (n 6) in the external solutionscontaining 1 mM Mg2�, 10 mM Mg2�, 0.5 mM Ca2�, 2mM Ca2�, and 10 mM Ca2�, respectively (see Fig. 6,A and C, but the entire outward currents were notshown). The increase in outward current by low pH waslarger in the external solutions with higher divalentconcentrations; and the effect of protons on outward

Figure 6. Protons increase TRPM7 inwardcurrents by competing with Ca2� and Mg2� forbinding sites. Current recorded in 2 mM Ca2�

Tyrode’s at pH 7.4 (red) is shown in A, C, E,and G as a reference. (A) Representativetraces of TRPM7 currents recorded in 0.5 and10 mM external Ca2� at pH 4.0. (B) Normal-ized increase in TRPM7 inward currents atindicated pH in the external solutions con-taining 0.5 and 10 mM Ca2�. The dotted linesrepresent the results obtained in 2 mM exter-nal Ca2� as shown in Fig 2. The pH values for50% maximal response were 5.1 � 0.2, 4.5 �0.5, and 3.4 � 0.1 in 0.5, 2, and 10 mM ex-ternal Ca2� (mean � SEM, n 6, 8, and 6),respectively. The Hill coefficient factors were1.0 for the dose–response curves in B, D, F,and H. (C) Representative TRPM7 currentselicited by voltage ramps in 1 and 10 mMexternal Mg2� at indicated pH values. (D)Normalized increase in TRPM7 inward currentamplitude at indicated pH in the externalsolutions containing various Mg2� concentra-tions. The pH values for 50% maximal re-sponse were pH 4.6 � 0.5 and pH 3.6 � 0.2in 1 and 10 mM external Mg2�, respectively(mean � SEM, n 6 for each group). (E)Representative currents elicited by voltageramps in 100 �M external Ca2� at pH 7.4 andpH 4.0. (F) Normalized current amplitude atpH 4.0 and pH 7.4 was plotted as a functionof external Ca2� concentrations. The Ca2�

concentration required to block 50% of themonovalent currents was 47.1 � 7.2 �M at pH7.4 and 5.6 � 0.7 mM at pH 4.0 (n 6). (G)Original recordings of TRPM7 in 9.9 and 10�M external Mg2� at indicated pH values. (H)Normalized current plotted as a function ofMg2� concentrations at pH 7.4 and pH 4.0.The IC50 was 5.9 � 1.2 mM at pH 4.0, and5.4 � 0.5 �M at pH 7.4 (mean � SEM, n 6), respectively.

that protons compete with Ca2� and Mg2� for the bind-ing sites in the TRPM7 pore.

We observed that protons produced different effectson outward currents under different external divalentcation conditions. For example, compared with thecurrents at pH 7.4 in 2 mM external Ca2�, the outwardcurrent amplitude at pH 4.0 measured at �100 mV waschanged by �41.1 � 5.4% (n 6), �276.7 � 35.3%

Jiang et al. 145

TRPM7 current was more pronounced in Mg2�-con-taining than in Ca2�-containing external solutions. Thisis presumably due to the fact that, at normal pH (7.4),Mg2� exhibits a stronger block on TRPM7 outward cur-rent than that of Ca2� (Kerschbaum et al., 2003; Mon-teilh-Zoller et al., 2003). Therefore, when the Mg2�

block is removed by protons, larger changes in TRPM7outward current are observed. These results are inagreement with the notion that protons compete withCa2� and Mg2� for binding sites in the TRPM7 pore.

If protons and divalent cations compete for bindingsites, and the binding of protons to the external sitesof the channel pore allows monovalent ions to passthrough TRPM7, one would expect that the inhibitoryeffects of Ca2� and Mg2� on monovalent currentsshould be influenced by proton concentrations. Wetherefore studied the inhibitory effects of Ca2� andMg2� on TRPM7 monovalent currents at pH 7.4 and pH4.0. At pH 4.0, the free Ca2� concentration was 7.7 �Mand the free Mg2� concentration was 9.9 �M, althoughthe same concentrations of EDTA and EGTA can de-

crease free Ca2� and Mg2� concentrations to �1 or 10nM at pH 7.4 (see MATERIALS AND METHODS). Fig.6 E shows representative recordings of TRPM7 in thepresence of 100 �M Ca2� at pH 4.0 and 7.4, respectively.IC50 was 47.1 �M Ca2� at pH 7.4 and 5.6 mM Ca2� at pH4.0. Similarly, 10 �M Mg2� produced more inhibitionon TRPM7 at pH 7.4 than that at pH 4.0 (Fig. 6 G). TheIC50 of Mg2� at pH 4.0 (5.9 mM) is about 1,000-fold dif-ferent from that at pH 7.4 (5.4 �M) (Fig. 6 H). The sig-nificantly decreased Ca2� and Mg2� affinities (Fig. 6, Fand H) to TRPM7 channels in the acidic external solu-tions further indicate that protons compete with diva-lent Ca2� and Mg2� for binding sites, thereby allowingmonovalent cations to pass through TRPM7 channels.

Anomalous-mole Fraction Behavior of H� PermeationAs protons compete with divalent cations for bindingsites in the channel pore, we tested whether protonspass through TRPM7 channels at low external divalentcation concentrations. No current was observed in theisotonic NMDG solution at pH 7.4, 7.0, or 6.0 (Fig. 7, Aand B). However, inward currents were observed in iso-tonic NMDG solutions at pH 5.0, 4.0, and 3.0. SinceNMDG is nonpermeant (Jiang et al., 2003; Kozak andCahalan, 2003), it seems that the inward current is car-ried by protons. This notion is supported by the factthat current amplitude increased with increasing pro-ton concentrations (Fig. 7 B). A high intracellular freeMg2� concentration (3 mM), which inactivates TRPM7channels, prevented the development of proton cur-rents in TRPM7-expressing cells (unpublished data),suggesting that the proton-carried current was medi-ated by TRPM7 channels. Fig. 7 (C and D) shows anom-alous mole fraction behavior of H� permeation. Thelargest current amplitude was observed at 10 �M extra-cellular Ca2� or Mg2�, whereas the smallest current am-plitude occurred at 10 mM extracellular Ca2� or Mg2�.These anomalous-mole fraction effects further indicatethat protons compete with Ca2� and Mg2� for bindingsites in the external pore of TRPM7, consistent with theresults shown in Fig. 6.

Effects of Protons on the Endogenous TRPM7-like Current MIC/MagNuM in RBL CellsEndogenous TRPM7-like currents MIC/MagNuM havebeen identified in a variety of cells (Nadler et al., 2001;Hermosura et al., 2002; Kozak et al., 2002; Prakriya andLewis, 2002; Runnels et al., 2002; Aarts et al., 2003; Jianget al., 2003). Since MIC/MagNuM in RBL cells havebeen well characterized (Hermosura et al., 2002; Kozaket al., 2002; Kozak and Cahalan, 2003, 2004), we choseto use RBL cells to study whether protons regulate en-dogenous MIC/MagNuM channels. As high concentra-tion of intracellular Ca2� blocks TRPM7 (Monteilh-Zoller et al., 2003), a pipette solution containing 15 nM

Figure 7. Currents carried by protons through TRPM7 channels.(A) Original recordings in isotonic NMDG solution at pH 7.4 andpH 4.0 using P3 pipette solution (note the inward current inducedat pH 4.0, red), and in 10 Ca2�/NMDG solution at pH 4.0 (green)in a HEK-293 cell overexpressing TRPM7. The proton current waslargely blocked by 10 mM Ca2�. (B) Averaged proton current atvarious test pH. Protons carry measurable currents at pH 5.0, andpronounced currents at pH 4.0 and 3.0 were observed (mean �SEM, n 6). *, P � 0.05; **, P � 0.01. (C and D) Anomalous molefraction behavior of H� permeation at pH 4.0. Current amplitudemeasured at �120 mV at indicated Ca2� or Mg2� concentrationwas normalized in reference to the current amplitude at 10 �Mextracellular Ca2� or Mg2� (mean � SEM, n 7).


Ca2� with weak buffering (1 mM EGTA) was used in thisexperiment. A low intracellular Ca2� buffering condi-tion (1 mM EGTA) was reported to be able to deactivateICRAC after its activation (Zweifach and Lewis, 1995) andwas used for recording of TRPM7-like MIC/MagNuMcurrents in Jurkat cells (Prakriya and Lewis, 2002). Wealso included 5 mM CsCl in the external solution toeliminate the endogenous potassium current. NaCl wasreplaced by NaSO3CH3 in the Tyrode’s solution at pH4.0 to prevent contamination from proton-activated Cl�

currents. After break-in, a voltage ramp protocol wasused to monitor current for 10 min to ensure the com-plete deactivation of ICRAC (Prakriya and Lewis, 2002)and full activation of TRPM7 (Kozak et al., 2002). Simi-lar to TRPM7 currents in the heterologous expressionsystem, the MIC/MagNum currents in RBL cells wereincreased dramatically when the pH of the external Ty-rode’s solution changed from 7.4 to 4.0 (Fig. 8 A). Theincrease elicited by low pH was not affected by adding200 �M amiloride (unpublished data), suggesting thatthe increase in MIC/MagNuM current was not due toproton-activated Na� channels. A best fit of normalizedconcentration-dependent increases in inward current(Fig. 8, B and C) yielded an EC50 of pH 4.2 � 0.4 (Fig.8 D), which is similar to the pH value required to in-duce 50% of maximal response in the heterologouslyexpressed TRPM7 currents (see Fig. 2 G). The similarresponse of TRPM7 and MIC/MagNuM to acidic pHprovides further evidence that MIC/MagNuM is en-coded by TRPM7.

D I S C U S S I O N

We have shown that low extracellular pH significantlyenhanced TRPM7 inward current by increasing the

monovalent cation permeability. The mechanism bywhich protons increase monovalent currents is likely tobe through competition with divalent cations for bind-ing sites in the external pore of TRPM7, thereby remov-ing the divalent cation block on the monovalent cur-rents. The pH sensitivity of TRPM7 and native MIC/MagNuM channels suggests that TRPM7 may play animportant role under acidic pathological conditions.

TRPM7 Is an Acid-sensitive Ion ChannelPrevious studies have shown that TRPM7 exhibits manyunique features (Nadler et al., 2001; Runnels et al.,2001; Monteilh-Zoller et al., 2003; Schmitz et al., 2003),including the observation that it conducts small inwardcurrent at physiological voltages (�100 to �40 mV)and large outward current at �100 mV, producingthe characteristic outwardly rectifying I-V. The presentstudy extends our understanding about TRPM7 byshowing that TRPM7 is also a pH-sensitive ion channel.We show that a decrease in external pH evokes amarked increase in TRPM7 inward currents, with an�10-fold increase at pH 4.0 and �1–2-fold increaseat pH 6.0 in the presence of a physiological externalCa2� concentration (2 mM). The dramatic increase inTRPM7 inward current by pH 4.0 transforms the nor-mally outwardly rectifying I-V curve to a double-rectify-ing I-V shape. We found that the increase in TRPM7current by protons was more pronounced when the ex-ternal Ca2� and Mg2� concentrations were decreased.Although the physiological function of TRPM7 is notcompletely understood, it has been shown that TRPM7plays an important role in neuronal cell death causedby anoxia (Aarts et al., 2003). Given that native TRPM7is only active to a small degree in the presence of physi-ological intracellular Mg2�, an approximately one- to

Figure 8. Effects of protons on MIC/MagNuM in RBLcells. Experiments were performed in external solutionswith NaCl replaced by NaSO3CH3. (A) Representative MICcurrents elicited by voltage ramps ranging from �120 to�100 mV in the external solutions at pH 7.4 (red) and 4.0(black) in a RBL cell. (B) Concentration-dependent effectsof protons on MIC inward currents in a representative RBLcell. The y axis was scaled to illustrate inward currents.(C) Inward current amplitude measured at �120 mV atdifferent test pH. External solution was changed to Tyrode’ssolution at pH 7.4 after each application of low pH externalsolution. (D) Concentration-dependent effects of protonson MIC currents. Best fit of the normalized change yieldedan EC50 4.2 � 0.4 (mean � SEM, n 6; Hill coefficientfactor was 1).

Jiang et al. 147

twofold increase in TRPM7 inward current at pH 6.0suggests that TRPM7 may play an important role underacidic pathological conditions in which extracellularpH may decrease to pH 6.0 (Jacobus et al., 1977;Stevens et al., 1991; Steen et al., 1992).

Potentiation of TRPM7 inward currents by externalprotons is a novel feature of TRPM7. A previous studyshowed that a decrease in intracellular pH (pHi) inhib-its monovalent Na� currents through MIC channels inJurkat T lymphocytes (Kerschbaum and Cahalan, 1998)and RBL cells (Braun et al., 2001). It is proposed thatinhibition of MIC or TRPM7 by internal Mg2�, otherpolyvalent cations, and H� represents a general electro-static cationic screening process (Kozak and Cahalan,2003, 2004). We did not study how intracellular low pHaffects TRPM7 in the present study, because the intra-cellular pH should not be changed due to the highconcentration of HEPES buffering under our experi-mental conditions. The marked increase in TRPM7 in-ward current by external acidic pH shown in thepresent study and the inhibitory effects on TRPM7 out-ward currents by pHi as previously reported (Kersch-baum and Cahalan, 1998; Braun et al., 2001; Kozak andCahalan, 2003, 2004) indicate that TRPM7 is an acid-sensitive ion channel.

Potential Mechanism by which Protons Potentiate TRPM7 Inward CurrentsWe showed that both Ca2� and Mg2� exhibit anoma-lous mole-fraction effects at normal physiological pH.The apparent affinity of TRPM7 is 47.1 �M for Ca2�

and 5.4 �M for Mg2�, similar to the previously reportedCa2� (20 �M, at �120 mV) (Fomina et al., 2000) andMg2� affinity (3 �M, at �120 mV) to native MIC/Mag-NuM channels (Kerschbaum et al., 2003). Under nor-mal physiological Ca2� (2 mM) and Mg2� (0.7–1.1 mM)concentrations (Konrad et al., 2004), we showed thatmonovalent cations contribute to the inward currentsof TRPM7 (Fig. 3), and the contribution of monovalentcurrents becomes more pronounced under acidic con-ditions (pH 4.0, Fig. 4).

Several lines of evidence shown in the present studyindicate that external protons increase TRPM7 inwardcurrents by competing with divalent cations for bindingsites in the TRPM7 pore, thereby enhancing monova-lent cation permeability. First, there was a concentra-tion-dependent increase in monovalent cation conduc-tance and reversal potential for Na�, K�, and Cs� at pH4.0, indicating that the enhanced inward TRPM7currents resulted from an increased monovalent cat-ion permeability. Second, the half-maximal pH waschanged toward acidic pH direction by 0.6 and 1.7 pHunits when the extracellular Ca2� was increased from0.5 to 2 and 10 mM, respectively; similarly, an increaseof external Mg2� concentration from 1 to 10 mM

shifted the half-maximal pH toward acidic pH directionby 1.0 pH unit (Fig. 6). The decreased proton affinityto TRPM7 at higher extracellular divalent concentra-tions indicate that protons compete with Ca2� andMg2� for the same binding sites, therefore, at higherconcentrations of divalent cations, more protons arerequired to induce 50% of maximal response. Third,high proton concentrations significantly decreased theaffinity of Ca2� and Mg2� to TRPM7. The Ca2� affinityto TRPM7 was decreased by �100-fold and the Mg2� af-finity was decreased by �1,000-fold when the externalpH was changed from 7.4 to 4.0. Fourth, anomalousmole fraction permeation of protons indicate that pro-tons compete with Ca2� and Mg2� for the binding sitesin the external pore of TRPM7 (Fig. 7). Taken to-gether, it seems that at physiological pH, TRPM7 onlypermeates a small inward current due to the Ca2� andMg2� block on the monovalent current; whereas atacidic pH, Ca2� and Mg2� are outcompeted by protons,which relieves the block on the monovalent currentand elicits a large inward current carried by monova-lent cations.

Where are the binding sites for divalent cations andprotons? A high affinity site for binding Mg2� withinthe electric field and two low-affinity sites have beenproposed for MIC channel by Kerschbaum et al. (2003)using the Eyring rate model. Kerschbaum et al. alsoproposed that the internal Mg2� inhibits MIC in avoltage-independent manner, suggesting that internalMg2� is unable to access the pore from the inside (Kersch-baum et al., 2003). Our results show that externalprotons elicit a marked increase in TRPM7 currentsat hyperpolarized potentials but only a small increaseat positive potentials. These voltage-dependent effectssuggest that protons can access the TRPM7 pore andcompete with Mg2� and Ca2� for binding. The anoma-lous mole fraction behavior of Ca2�, Mg2� (Fig. 3 F),and H� permeation (Fig. 7, C and D) indicates thatthey can bind to the TRPM7 pore, and compete forbinding sites within the pore (Fig. 6).

Proton competition for Ca2� binding sites and theconsequent channel opening has been proposed as thegating mechanism for ASIC3 (Immke and McCleskey,2001, 2003), a proton-activated Na� channel of thedegenerin family (Waldmann and Lazdunski, 1998;Immke and McCleskey, 2003). Immke and McCleskey(2003) showed that the Ca2� affinity was changed fromKd 12 �M at pH 7.4 to Kd 100 �M at pH 7.0, suchthat Ca2� is released from a binding site and Na� canpass through the channel at millimolar Ca2� concentra-tions. The authors predict that, like the Ca2� chelatorEGTA (with four titratable acid groups), the titratableCa2� binding site of ASIC3 is able to bind four protons(Hill coefficient is 4). Our data suggest that protons en-hance TRPM7 current by a similar competing mecha-


nism, such that binding of protons to TRPM7 relievesthe blockade of Ca2� and Mg2�, thereby allowing Na�

to pass through TRPM7. Since the dose–responsecurves can be best fitted by sigmoidal does–responseequation (see MATERIALS AND METHODS) with theHill coefficient factor of 1, we assume that there isprobably one binding site. We do not know if the bind-ing site is a specific amino acid residue or a site formedby several amino acid residues. It was reported that pro-tonation of voltage-gated Ca2� channels requires multi-ple carboxylates of the glutamic acid (Glu) residues toform a single high affinity site (Chen and Tsien, 1997).The Glu residues are also important for proton regula-tion on TRPV1 (Tominaga et al., 1998; Jordt et al.,2000; Ryu et al., 2003) and TRPV5 (Yeh et al., 2003).There are seven negatively charged Glu and Asparticacid (Asp) residues between transmembrane domain 5(TM5) and TM6 of TRPM7, which may be involved inCa2�, Mg2�, or H� binding. The half-maximum pH(pH 4.5) for TRPM7 is also close to the pKa of free Glu(�pH 4.0) and Asp (�pH 3.8). Thus, it is possible thatGlu or Asp in the TRPM7 pore may serve as the bind-ing sites for external Ca2� and Mg2�, and are also ableto bind to protons, so that monovalent cations canreadily pass through when the binding sites are occu-pied by protons. Alternatively, instead of competitivebinding, protons may titrate away the block of Ca2� andMg2� on monovalent current by causing conforma-tional changes. Further studies are required to eluci-date the proton binding sites and detailed mechanismsby which protons increase TRPM7 inward monovalentcurrents.

Potential SignificanceIt has been suggested that TRPM7 plays an importantphysiological role in Mg2� homeostasis, neuronal celldeath, and cell viability (Nadler et al., 2001; Schling-mann et al., 2002; Walder et al., 2002; Aarts et al., 2003;Schmitz et al., 2003; Chubanov et al., 2004). We dem-onstrated here that acidic pH 6.0 increases TRPM7 in-ward current by approximately one- to twofold. Suchan acidic condition (pH � 6) can occur during variousforms of tissue injury (Jacobus et al., 1977; Stevens etal., 1991; Steen et al., 1992), or during repetitive nerveactivities, ischemia, and seizures (Siesjo, 1988; Cheslerand Kaila, 1992), suggesting that TRPM7 may play arole under acidic pathological conditions. However,without knowing the real physiological functions ofTRPM7, it is difficult to predict a potential role ofTRPM7 under acidic conditions. We have investigatedif there are other factors that may change the pH sensi-tivity of TRPM7 closer to physiological pH. With 20 �MPIP2 in the pipette solution, the EC50 was pH 4.7 � 0.7(mean � SEM, n 6, Hill coefficient 1.0; unpublisheddata), which is not significantly different from the EC50

obtained without PIP2 in the pipette solution (Fig. 2).In addition, with normal physiological intracellularMg2� concentrations in a native system, the MIC/Mag-NuM channel is only active to a small degree. Furtherstudies are required to elucidate the physiological andpathological roles of TRPM7 under normal and acidicconditions.

Like TRPM7, the native MIC/MagNuM current inRBL cells showed a similar acidic potentiation to thatseen for TRPM7 currents in the heterologous expres-sion system (Fig. 8), indicating that MIC/MagNuM isencoded by TRPM7. A recent study by Gwanyanya et al.(2004) showed that MIC in cardiac myocytes and RBLcells were inhibited by acidic pH. It is not clear whythere is a discrepancy between our and their results.One difference is that they evaluated pH effects onMIC in divalent-free solutions, whereas we used physio-logical external solutions containing divalent cations inthe present study.

Protons have been reported to regulate channel ac-tivities in different channel superfamilies (Hille, 2003;Holzer, 2003), including TRP channel superfamily. Ithas been shown that TRPV1 and TRPV4 are enhancedby low pH, whereas TRPV5 (Yeh et al., 2003) andTRPM5 (Liu et al., 2005) are inhibited by protons. Weshow here that protons dramatically potentiate TRPM7inward current by competing with divalent cations forbinding sites. The pH sensitivity is a novel feature ofTRPM7, and the results in the present study not onlyprovide a clue as to the potential functions of TRPM7in vivo, but also may help to identify the amino acid res-idues that are important in the ion selectivity of TRPM7channels.

Potential LimitationsProton permeation has been observed for voltage-gated Na� channels when the external solution is freeof Na� (Mozhayeva and Naumov, 1983; Hille, 2003).We showed anomalous mole fraction behavior of H�

permeation in the external solutions free of permeantmonovalent cations (Fig. 7). In 2 mM Ca2�/NMDG/pH 4.0 solution, the proton carried current is 4.4 pA/pF, corresponding to 17% of the current amplitude ob-tained in 10 �M Ca2�/NMDG/pH 4.0 (Fig. 7 C). Thisproton current amplitude is similar to the value shownin Fig. 5 B (6.5 pA/pF, pH 4.0), indicating that the cur-rent at pH 4.0 (Fig. 5 B) is mainly carried by protons,and consistent with the notion that protons competewith divalent cations and therefore almost blocked allthe divalent cation current under the conditions shownin Fig. 5 (A and B).

It is possible that in the normal Tyrode’s solution atpH 4.0, protons may pass through TRPM7 along withNa� or K�. If this were the case, proton-carried currentmay have contributed to the inward monovalent cur-

Jiang et al. 149

rents at pH 4.0. However, given that the current ampli-tude (6.5 pA/pF, Fig. 5 B) in NMDG solution at pH 4.0is only �2.5% of the inward current amplitude (247pA/pF, Fig. 2) obtained in normal Tyrode’s solutions atpH 4.0, contribution of proton-carried inward currentsin the normal Tyrode’s solutions at pH 4.0 should be�2.5%, and should not contaminate the experimentalresults.

At pH 3.0, the I-V curve of TRPM7 elicited by voltageramps seems different from those at pH �3.0 (Fig. 2E), and the TRPM7 outward currents were inhibited bypH 3.0. In addition, a strong inactivation or desensitiza-tion was observed (Fig. 2 F) when cells were continu-ously exposed to the external solution at pH 3.0. Thefollowing potential mechanisms may account for theabove observations. First, TRPM7 is desensitized at pH3.0; second, protons may pass through TRPM7, result-ing in low intracellular pH, which inhibits TRPM7outward current (Kozak and Cahalan, 2004); andthird, protons may exhibit complex effects on TRPM7.Further studies are required to reveal the detailedmechanisms.

ConclusionIn conclusion, we have demonstrated that acidic pHsignificantly increases TRPM7 inward monovalent cur-rents by competing with divalent cations for bindingsites. The pH sensitivity represents a novel feature ofTRPM7. We showed that MIC/MagNuM currents weresimilarly potentiated by protons, suggesting that MIC/MagNuM is encoded by TRPM7. The large TRPM7 in-ward current elicited by low pH suggests that TRPM7may play a role under acidic pathological conditions.Further studies are required to elucidate the mecha-nism by which protons potentiate TRPM7, as well as thepotential significance of TRPM7 under acidic patholog-ical conditions.

We thank Dr. Andrew Scharenberg for providing the HEK-293cells stably expressing TRPM7. We thank Dr. Alan Fein for con-structive suggestions and comments. And we’d like to thankDavid Clapham, Bruce Liang, Kim Dodge, Haoxin Xu, LorenRunnels, Laurinda Jaffe, and Dejian Ren for helpful comments.

This work was supported in part by an award from the Ameri-can Heart Association and by a National Institutes of Healthgrant R01 HL078960-01A1 to L. Yue.

Olaf S. Andersen served as editor.

Submitted: 22 September 2004Accepted: 10 June 2005

R E F E R E N C E SAarts, M., K. Iihara, W.L. Wei, Z.G. Xiong, M. Arundine, W. Cerwin-

ski, J.F. MacDonald, and M. Tymianski. 2003. A key role forTRPM7 channels in anoxic neuronal death. Cell. 115:863–877.

Askwith, C.C., J.A. Wemmie, M.P. Price, T. Rokhlina, and M.J.Welsh. 2004. Acid-sensing ion channel 2 (ASIC2) modulatesASIC1 H�-activated currents in hippocampal neurons. J. Biol.

Chem. 279:18296–18305.Braun, F.J., L.M. Broad, D.L. Armstrong, and J.W. Putney Jr. 2001.

Stable activation of single Ca2� release-activated Ca2� channels indivalent cation-free solutions. J. Biol. Chem. 276:1063–1070.

Chen, X.-H., and R.W. Tsien. 1997. Aspartate substitutions establishthe concerted action of P-region glutamates in repeats I and IIIin forming the protonation site of L-type Ca2� channels. J. Biol.Chem. 272:30002–30008.

Cherny, V.V., L.M. Henderson, and T.E. DeCoursey. 1997. Protonand chloride currents in Chinese hamster ovary cells. Membr. CellBiol. 11:337–347.

Chesler, M., and K. Kaila. 1992. Modulation of pH by neuronal ac-tivity. Trends Neurosci. 15:396–402.

Chubanov, V., S. Waldegger, M. Mederos y Schnitzler, H. Vitzthum,M.C. Sassen, H.W. Seyberth, M. Konrad, and T. Gudermann.2004. Disruption of TRPM6/TRPM7 complex formation by amutation in the TRPM6 gene causes hypomagnesemia with sec-ondary hypocalcemia. Proc. Natl. Acad. Sci. USA. 101:2894–2899.

Clapham, D.E. 2003. TRP channels as cellular sensors. Nature. 426:517–524.

Fleig, A., and R. Penner. 2004. Emerging roles of TRPM channels.Novartis Found. Symp. 258:248–258; discussion 258–266.

Fomina, A.F., C.M. Fanger, J.A. Kozak, and M.D. Cahalan. 2000.Single channel properties and regulated expression of Ca2� re-lease-activated Ca2� (CRAC) channels in human T cells. J. CellBiol. 150:1435–1444.

Gwanyanya, A., B. Amuzescu, S.I. Zakharov, R. Macianskiene, K.R.Sipido, V.M. Bolotina, J. Vereecke, K. Mubagwa. 2004. Magne-sium-inhibited, TRPM6/7-like channel in cardiac myocytes: per-meation of divalent cations and pH-mediated regulation. J. Phys-iol. 559:761–776.

Hanano, T., Y. Hara, J. Shi, H. Morita, C. Umebayashi, E. Mori, H.Sumimoto, Y. Ito, Y. Mori, and R. Inoue. 2004. Involvement ofTRPM7 in cell growth as a spontaneously activated Ca2� entrypathway in human retinoblastoma cells. J. Pharmacol. Sci. 95:403–419.

Harteneck, C., T.D. Plant, and G. Schultz. 2000. From worm toman: three subfamilies of TRP channels. Trends Neurosci. 23:159–166.

Hermosura, M.C., M.K. Monteilh-Zoller, A.M. Scharenberg, R. Pen-ner, and A. Fleig. 2002. Dissociation of the store-operated cal-cium current I(CRAC) and the Mg-nucleotide-regulated metalion current MagNuM. J. Physiol. 539:445–458.

Hille, B. 2003. Ion Channels of Excitable Membranes. Third edi-tion. Sinauer Associates, Inc. Sunderland, MA. 471–502.

Holzer, P. 2003. Acid-sensitive ion channels in gastrointestinal func-tion. Curr. Opin. Pharmacol. 3:618–625.

Immke, D.C., and E.W. McCleskey. 2001. Lactate enhances theacid-sensing Na� channel on ischemia-sensing neurons. Nat. Neu-rosci. 4:869–870.

Immke, D.C., and E.W. McCleskey. 2003. Protons open acid-sensingion channels by catalyzing relief of Ca2� blockade. Neuron. 37:75–84.

Jacobus, W.E., G.J. Taylor, D.P. Hollis, and R.L. Nunnally. 1977.Phosphorus nuclear magnetic resonance of perfused working rathearts. Nature. 265:756–758.

Jiang, X., E.W. Newell, and L.C. Schlichter. 2003. Regulation of aTRPM7-like current in rat brain microglia. J. Biol. Chem. 278:42867–42876.

Jordt, S.E., M. Tominaga, and D. Julius. 2000. Acid potentiation ofthe capsaicin receptor determined by a key extracellular site.Proc. Natl. Acad. Sci. USA. 97:8134–8139.

Kerschbaum, H.H., and M.D. Cahalan. 1998. Monovalent perme-ability, rectification, and ionic block of store-operated calciumchannels in Jurkat T lymphocytes. J. Gen. Physiol. 111:521–537.


Kerschbaum, H.H., J.A. Kozak, and M.D. Cahalan. 2003. Polyvalentcations as permeant probes of MIC and TRPM7 pores. Biophys. J.84:2293–2305.

Konrad, M., K.P. Schlingmann, and T. Gudermann. 2004. Insightsinto the molecular nature of magnesium homeostasis. Am. J.Physiol. Renal Physiol. 286:F599–F605.

Kozak, J.A., and M.D. Cahalan. 2003. MIC channels are inhibitedby internal divalent cations but not ATP. Biophys. J. 84:922–927.

Kozak, J.A., and M.D. Cahalan. 2004. Regulation of MIC andTRPM7 channels by internal polyvalent cations. Biophys J. Ab-stract.

Kozak, J.A., H.H. Kerschbaum, and M.D. Cahalan. 2002. Distinctproperties of CRAC and MIC channels in RBL cells. J. Gen. Phys-iol. 120:221–235.

Liu, D., Z. Zhang, and E.R. Liman. 2005. Extracellular acid blockand acid-enhanced inactivation of the Ca2�-activated cationchannel TRPM5 involve residues in the S3-S4 and S5-S6 extracel-lular domains. J. Biol. Chem. 280:20691–20699.

Monteilh-Zoller, M.K., M.C. Hermosura, M.J. Nadler, A.M. Scharen-berg, R. Penner, and A. Fleig. 2003. TRPM7 provides an ionchannel mechanism for cellular entry of trace metal ions. J. Gen.Physiol. 121:49–60.

Montell, C. 2001. Physiology, phylogeny, and functions of the TRPsuperfamily of cation channels. Sci STKE. 2001:RE1.

Mozhayeva, G.N., and A.P. Naumov. 1983. The permeability of so-dium channels to hydrogen ions in nerve fibres. Pflugers Arch.396:163–173.

Nadler, M.J., M.C. Hermosura, K. Inabe, A.L. Perraud, Q. Zhu, A.J.Stokes, T. Kurosaki, J.P. Kinet, R. Penner, A.M. Scharenberg, andA. Fleig. 2001. LTRPC7 is a Mg.ATP-regulated divalent cationchannel required for cell viability. Nature. 411:590–595.

Prakriya, M., and R.S. Lewis. 2002. Separation and characterizationof currents through store-operated CRAC channels and Mg2�-inhibited cation (MIC) channels. J. Gen. Physiol. 119:487–507.

Runnels, L.W., L. Yue, and D.E. Clapham. 2001. TRP-PLIK, a bi-functional protein with kinase and ion channel activities. Science.291:1043–1047.

Runnels, L.W., L. Yue, and D.E. Clapham. 2002. The TRPM7 chan-nel is inactivated by PIP(2) hydrolysis. Nat. Cell Biol. 4:329–336.

Ryazanova, L.V., M.V. Dorovkov, A. Ansari, and A.G. Ryazanov.2004. Characterization of the protein kinase activity of TRPM7/ChaK1, a protein kinase fused to TRP ion channel. J. Biol. Chem.279:3708–3716.

Ryu, S., B. Liu, and F. Qin. 2003. Low pH potentiates both capsai-cin binding and channel gating of VR1 receptors. J. Gen. Physiol.122:45–61.

Schlingmann, K.P., S. Weber, M. Peters, L. Niemann Nejsum, H.Vitzthum, K. Klingel, M. Kratz, E. Haddad, E. Ristoff, D. Dinour,et al. 2002. Hypomagnesemia with secondary hypocalcemia iscaused by mutations in TRPM6, a new member of the TRPMgene family. Nat. Genet. 31:166–170.

Schmitz, C., A.L. Perraud, C.O. Johnson, K. Inabe, M.K. Smith, R.Penner, T. Kurosaki, A. Fleig, and A.M. Scharenberg. 2003. Regu-

lation of vertebrate cellular Mg2� homeostasis by TRPM7. Cell.114:191–200.

Siesjo, B.K. 1988. Acidosis and ischemic brain damage. Neurochem.Pathol. 9:31–88.

Steen, K.H., P.W. Reeh, F. Anton, and H.O. Handwerker. 1992. Pro-tons selectively induce lasting excitation and sensitization to me-chanical stimulation of nociceptors in rat skin, in vitro. J. Neuro-sci. 12:86–95.

Stevens, C.R., R.B. Williams, A.J. Farrell, and D.R. Blake. 1991. Hyp-oxia and inflammatory synovitis: observations and speculation.Ann. Rheum. Dis. 50:124–132.

Takezawa, R., C. Schmitz, P. Demeuse, A.M. Scharenberg, R. Pen-ner, and A. Fleig. 2004. Receptor-mediated regulation of theTRPM7 channel through its endogenous protein kinase domain.Proc. Natl. Acad. Sci. USA. 101:6009–6014.

Tominaga, M., M.J. Caterina, A.B. Malmberg, T.A. Rosen, H. Gil-bert, K. Skinner, B.E. Raumann, A.I. Basbaum, and D. Julius.1998. The cloned capsaicin receptor integrates multiple pain-producing stimuli. Neuron. 21:531–543.

Vennekens, R., J.G.J. Hoenderop, J. Prenen, M. Stuiver, P.H.G.M.Willems, G. Droogmans, B. Nilius, and R.J.M. Bindels. 2000. Per-meation and gating properties of the novel epithelial Ca2� chan-nel. J. Biol. Chem. 275:3963–3969.

Walder, R.Y., D. Landau, P. Meyer, H. Shalev, M. Tsolia, Z. Boro-chowitz, M.B. Boettger, G.E. Beck, R.K. Englehardt, R. Carmi,and V.C. Sheffield. 2002. Mutation of TRPM6 causes familial hy-pomagnesemia with secondary hypocalcemia. Nat. Genet. 31:171–174.

Waldmann, R., and M. Lazdunski. 1998. H�-gated cation channels:neuronal acid sensors in the NaC/DEG family of ion channels.Curr. Opin. Neurobiol. 8:418–424.

Yeh, B.-I., T.-J. Sun, J.Z. Lee, H.-H. Chen, and C.-L. Huang. 2003.Mechanism and molecular determinant for regulation of rabbittransient receptor potential type 5 (TRPV5) channel by extracel-lular pH. J. Biol. Chem. 278:51044–51052.

Yermolaieva, O., A.S. Leonard, M.K. Schnizler, F.M. Abboud, andM.J. Welsh. 2004. Extracellular acidosis increases neuronal cellcalcium by activating acid-sensing ion channel 1a. Proc. Natl.Acad. Sci. USA. 101:6752–6757.

Yue, L., J.L. Feng, Z. Wang, and S. Nattel. 2000. Effects of ambasil-ide, quinidine, flecainide and verapamil on ultra-rapid delayedrectifier potassium currents in canine atrial myocytes. Cardiovasc.Res. 46:151–161.

Yue, L., B. Navarro, D. Ren, A. Ramos, and D. Clapham. 2002. Thecation selectivity filter of the bacterial sodium channel, NaCh-Bac. J. Gen. Physiol. 160:845–853.

Yue, L., J.B. Peng, M.A. Hediger, and D.E. Clapham. 2001. CaT1manifests the pore properties of the calcium-release-activatedcalcium channel. Nature. 410:705–709.

Zweifach, A., and R.S. Lewis. 1995. Slow calcium-dependent inacti-vation of depletion-activated calcium current. Store-dependentand -independent mechanisms. J. Biol. Chem. 270:14445–14451.

potentiation of trpm7 inward currents by...

Documents